JOURNAL OF PLANKTON RESEARCH
j
VOLUME
26
j
NUMBER
9
j
PAGES
1083–1093
j
2004
Seasonal variability and inter-specific differences in hatching of calanoid copepod resting eggs from sediments of the German Bight (North Sea) M. ENGEL* AND H.-J. HIRCHE ALFRED WEGENER INSTITUTE FOR POLAR AND MARINE RESEARCH, COLUMBUSSTRASSE, *CORRESPONDING AUTHOR:
27568 BREMERHAVEN, GERMANY
[email protected]
Received February 16, 2004; accepted in principle April 8, 2004; accepted for publication May 26, 2004; published online June 3, 2004
Hatching experiments were conducted on sediment samples collected on seven cruises between March 2002 and February 2003, at four stations near Helgoland (54110 N, 7 530 E). Samples were incubated for up to 12 months and examined weekly for calanoid copepod nauplii. 12807 nauplii hatched in total. Of these 36.43% were stage N1 (pooled), 44.08% were Temora longicornis (stage N2 and older), and 17.96% were Centropages hamatus (stage N2 and older). Some Acartia spp. and an unidentified species were found, too. Hatching varied significantly between sampling dates. At three stations, counts of all nauplii peaked in samples taken in April, while at one station the maximum was recorded in the sample collected in October. The seasonal pattern of hatching was similar in T. longicornis and C. hamatus. In March, April, October and February numbers of all nauplii were initially low, but increased over the following 2–3 weeks. In June, August and December, however, average numbers were comparatively high at first, but declined thereafter. In three of the four samples that were incubated for 12 months nauplii hatched from the sediment for the entire experimental period. Temora longicornis and C. hamatus displayed clearly distinct patterns of hatching in these long-term incubations.
INTRODUCTION To date resting eggs have been reported for 49 species of marine and estuarine calanoid copepods (Champeau, 1970; Belmonte and Puce, 1994; Marcus, 1996; Naess, 1996; Belmonte, 1997; Chen and Marcus, 1997; Guerrero and Rodriguez, 1998; Newton and Mitchell, 1999; Hall and Burns, 2001). They can remain viable in the sediment for many months or even years (Viitasalo, 1992; Katajisto, 1996). As resting eggs are frequently found in species that are absent from the water column for portions of the year, they are often regarded as a means of securing the survival of a population in times of unfavourable environmental conditions in the plankton. Other purposes that have been suggested include the temporal partitioning of the environment (Marcus, 1984), the prevention of overcrowding (Uye, 1980), and even the slowdown of the rate of evolutionary change (Hairston and De Stasio, 1988). Three types of
resting eggs have been distinguished: quiescent subitaneous, diapause and delayed hatching eggs (Chen and Marcus, 1997), which differ in their mode of development. Subitaneous eggs hatch within a few days, but can become quiescent in response to adverse environmental conditions. As soon as these improve, eggs are capable of hatching. In contrast, diapause eggs hatch only after the completion of a refractory phase, which may last several months, even if conditions are beneficial (Grice and Marcus, 1981). During this period development and/or metabolic processes apparently cease or drop to undetectable levels (Marcus, 1996). Unlike post-refractory diapause eggs, which may hatch soon after they perceive the appropriate environmental cues, nauplii emerge from delayed-hatching eggs only gradually over an extended period of time (Chen and Marcus, 1997). Due to their abundance, up to several million have been found per m2 (Marcus, 1984; Guerrero and Rodriguez,
doi:10.1093/plankt/fbh099, available online at www.plankt.oupjournals.org Journal of Plankton Research Vol. 26 No. 9 Ó Oxford University Press 2004; all rights reserved
JOURNAL OF PLANKTON RESEARCH
j
VOLUME
1998), resting eggs constitute an important component of bentho-pelagic coupling. Benthic processes and environmental conditions in the sediment, for example, influence survival and hatching rate of resting eggs and thereby affect the reproductive success of a species (Marcus and Lutz, 1998). As climate may act differently (spatial, temporal) on the benthic and pelagic systems, population dynamics of those copepod species that have resting eggs should have two control components, a benthic and a pelagic. Consequently, understanding the inter-annual variability of copepod populations in relation to climatology makes understanding egg bank dynamics a prerequisite. In surveys conducted in the 1980s, nauplii of Temora longicornis, Centropages hamatus, Centropages typicus, Labidocera wollastoni and Acartia spp. hatched from sediment samples collected from various regions around the British Isles, including the southern North Sea and the English Channel (Lindley, 1986, 1990). Taking into account the distribution of planktonic stages, Lindley concluded that water depth and bottom stress from tidal currents determine the abundance of resting eggs and hence the distributions of their pelagic populations, particularly in those species that overwinter in the form of resting eggs. Using these two physical parameters he predicted copepod eggs to be abundant in bottom sediments
26
j
NUMBER
9
j
PAGES
1083–1093
j
2004
in many areas around the British Isles including the German Bight (Lindley and Hunt, 1989; Lindley, 1990). Most of the species mentioned above are indeed important components of the zooplankton community in the German Bight. However, little is known about the overwintering strategies of copepods in this area of the North Sea, and the role of resting eggs in their seasonal dynamics has never been studied. Species like C. hamatus and C. typicus occur only seasonally in the plankton (Halsband and Hirche, 2001), but it is unknown whether they are advected to this area year after year or spend the winter as resting eggs in the sediment. It was therefore the aim here to study seasonal distribution and species composition of resting eggs in the German Bight, to follow the seasonal dynamics of hatching, and to study temporal patterns of hatching during long-term incubations over 1 year.
METHOD Sampling and experimental procedures Sediment cores were taken at four stations (Fig. 1) in the German Bight, North Sea. Water depth ranged from 22 to 42 m and sediments were either muddy sand or sand
Fig. 1. Map of the German Bight showing the four sampling stations.
1084
M. ENGEL AND H.-J. HIRCHE
j
HATCHING OF CALANOID COPEPOD RESTING EGGS
Table I: Water depth (m) and sediment composition (analysis was performed on an additional set of samples collected on 22 April 2002) at the four sampling stations Station
Water depth
size 63 mm) to obtain the sand fraction. Silt and clay were separated using the Atterberg technique (Mu¨ller, 1967), which is based on different sinking velocities of particles from different size classes.
Statistical analysis
Sediment composition
To compare results among sampling dates, the sum of all nauplii that were found in the first six screenings was calculated for every sample. Friedman’s two-way analysis of variance by ranks was used to detect an effect of this factor on the results in general. Subsequently, Wilcoxon’s matched pairs test was employed to check for significant differences between pairs of sampling dates (Sachs, 1978).
(m) Sand
Silt
Clay
(2000 – 63 mm)
(63 – 2 mm)
(< 2 mm)
1
22
97.50%
1.27%
1.23%
2
30
71.33%
17.51%
11.16%
3
42
74.23%
15.50%
10.27%
4
35
76.33%
12.50%
11.17%
RESULTS (Table I). All stations were sampled seven times between March 2002 and February 2003. Sampling was accomplished using a minicorer or by taking sediment cores from a box corer or a van Veen grab (Table II). For all three methods Perspex tubes with an inner diameter of 56 mm were used (i.e. a sample represents 24.6 cm2 of seafloor). Temperature of the overlying water was measured immediately after collection. Subsequently, the sediment core was carefully pushed upwards in the tube using a piston, and its top 5–7 cm (125–175 cm3) were spooned into a transparent 500 ml Kautex bottle. The bottle was topped up with 55 mm filtered seawater. For the rest of the cruise and during the experiments, the samples were kept at a temperature (Table II) close to that measured at their collection. In the laboratory samples collected on 22 April 2002 were incubated for 52 weeks (long-term), while the remaining 24 samples were incubated for 6–13 weeks (short-term). The light regime was light:dark 12:12 h in all experiments. The overlying water in the Kautex bottles was carefully poured off weekly over a 55 mm sieve and the bottles were refilled with 55 mm filtered seawater. Material retained by the sieve was screened for copepod nauplii under a dissecting microscope (Leica Wild M 10). Specimens of naupliar stage 2 (N2) and older were identified to the lowest possible taxonomic level, but nauplii of stage 1 (N1) were pooled, as they could not be reliably assigned to species. The 28 samples were screened 420 times in total (Table II).
Particle size distribution On 22 April 2002, an additional sediment sample was taken at each station for the analysis of particle size distribution (Table I). In the laboratory, 5 cm3 of sediment were suspended in 100 ml water and 100 ml 10% H2O2 was added to dissolve any organic material in the sample. Subsequently samples were sieved (mesh
Species and stage composition Calanoid copepod nauplii hatched from all 28 samples, but numbers varied widely, between 1 and 1155 per sample in screenings 1–6 (Table II). In 420 screenings 12 807 nauplii were found in total. They were assigned to five different groups: T. longicornis (N2 and older), C. hamatus (N2 and older), Acartia spp. (N2 and older), nauplii of an unidentified species, and pooled N1 (Table III). Abundance maxima in discrete samples were equivalent to 506 911 T. longicornis nauplii per m2 (October, station 1, 13 screenings) and 286 585 C. hamatus nauplii per m2 (April, station 4, 52 screenings). However, as these two species accounted for the vast majority of all specimens that were N2 and older, it is assumed that they also represent those classified as N1. Thus maximum abundance of T. longicornis and C. hamatus should even be higher. Since many pelagic calanoid copepods do not start to feed until they have developed to N3, speciesspecific differences in nauplii survival were not expected. By far the most frequent stage was N2 followed by N1, while older stages were extremely rare (Table III). The unidentified nauplius is laterally compressed, hunchbacked and, excluding the two furcal appendages, 160 mm in length. These appendages are of equal length and their bases, twice as long as the protruding parts, are clearly visible inside the animal’s abdomen. The shape of the nauplius suggests that it belongs to the calanoida rather than to the cyclopoida or harpacticoida, which are usually dorsoventrally depressed (Dussart and Defaye, 1995).
Seasonal variability Counts of nauplii also varied in between sampling dates. All nauplii that were found in screenings 1–6 in the four
1085
Table II: Sampling gear and incubation temperature is given for every sample, together with numbers of T. longicornis, C. hamatus, Acartia spp., the unidentified species, N1, and the total number of nauplii that were found to hatch from each sample in the first six screenings (s. 1–6) and in all screenings (all s.) the sample was subjected to (as indicated in column 5) Station
Gear
Temperature
Screenings
T. longicornis
C. hamatus
(s. 1–6)
(s. 1–6)
Acartia spp.
Unidentified sp.
N1
(s. 1–6)
(s. 1–6)
(s. 1–6)
Total
( C)
date
(all s.)
(all s.)
(all s.)
(all s.)
(all s.)
(s. 1–6)
(all s.)
11–13
1
vVG
5
6
2
2
1
1
0
0
0
0
2
2
5
5
Mar
2
MIC
5
6
49
49
10
10
0
0
0
0
24
24
83
83 124
2002
5
6
67
67
2
2
0
0
1
1
54
54
124
5
6
120
120
18
18
0
0
0
0
55
55
193
193
1
vVG
8
52
78
146
40
79
1
1
2
3
126
186
247
415
Apr
2
MIC
8
52
70
436
23
311
0
3
0
3
98
785
191
1538
2002
3
MIC
8
52
153
735
25
140
7
21
0
11
138
746
323
1653
4
MIC
8
52
349
1126
128
705
1
3
1
30
342
1297
821
3161
17
1
BC
16
9
0
0
2
2
0
0
0
0
0
0
2
2
Jun
2
MIC
16
9
0
9
1
1
0
0
0
0
0
4
1
14
22
9
2
16
1
3
0
0
0
0
7
12
10
31
9
33
38
45
46
14
14
0
0
8
10
100
108
21
1
BC
18
8
37
55
21
23
7
7
1
1
20
34
86
120
Aug
2
MIC
18
8
19
41
17
50
9
9
1
3
28
71
74
174
2002
3
MIC
18
8
65
97
34
42
5
6
0
0
21
29
125
174
4
MIC
18
8
81
141
43
63
11
11
0
9
28
42
163
266
1
BC
14
13
710
1247
278
487
0
0
1
3
166
329
1155
2066
15
9
16 16
NUMBER
MIC MIC
j
3 4
26
1086
2002
VOLUME
MIC MIC
j
3 4
JOURNAL OF PLANKTON RESEARCH
Sampling
j
MIC
14
13
79
191
15
39
0
0
0
0
22
66
116
296
3
MIC
14
13
39
41
39
43
1
5
0
0
61
93
140
182
4
MIC
14
13
80
243
8
33
1
1
1
1
43
91
133
369
1
BC
8
8
49
51
17
17
1
1
3
3
94
98
164
170
12–13 Dec
2
BC
8
8
2
23
4
9
0
0
0
0
8
34
14
66
2002
3
BC
8
8
8
15
2
2
0
0
2
2
5
30
17
49
4
BC
8
8
11
85
0
9
0
0
0
1
17
38
28
133
j
6
1
BC
5
9
429
529
115
143
0
0
24
32
301
341
869
1045
Feb
2
MIC
5
9
27
29
2
2
0
0
0
0
43
45
72
76
2004
1083–1093
2
2002
PAGES
Oct
2003
3
MIC
5
9
13
16
1
2
0
0
1
1
31
32
46
51
4
MIC
5
9
87
97
17
18
0
0
10
10
106
118
220
243
VVG, van Veen grab; MIC, minicorer; BC, boxcorer.
M. ENGEL AND H.-J. HIRCHE
j
HATCHING OF CALANOID COPEPOD RESTING EGGS
Table III: Species and stage composition (%) of all calanoid copepod nauplii found in this study
Table IV: The outcome from Wilcoxon’s matched pairs test Dec 02: Aug 02: Mar 02: Feb 03: Oct 02: Apr 02:
Total abundance (%) Jun 02: 5
Species Species composition
T. longicornis
44.08
Dec 02: 9
(12807 nauplii)
C. hamatus
17.96
Aug 02: 16
Acartia spp.
0.64
Mar 02: 16
0.89
Feb 03: 18
36.43
Oct 02: 22
Unidentified sp. N1 Stage Stage composition
N1
36.76
(12693 nauplii,
N2
61.25
unidentified sp.
N3
1.83
excluded)
N4
0.15
N5
Number of specimens
1500
1000
500
0
A
M
J
J
A
S
O
2002
16
18
4
11
11
13
17*
21**
7
7
9
13
17*
0
22
26
2
6
10
2
6
10
4
8 4
specimens of Acartia spp. and the unidentified sp. to compare their seasonal cycle. Considering each station separately gives a more detailed picture of seasonal variability. Patterns for N1 and all nauplii were almost identical at stations 2, 3 and 4. Pronounced maxima were recorded in April (Fig. 3a, d). Values accounted for 43.5–57.1% and 34.7–49.5% of the sum of N1 and of all nauplii that were found in the first six screenings in all samples collected at the corresponding stations, respectively. The maxima at station 1 were found in the samples taken in February (N1, 42.4%) and October (all nauplii, 45.6%). Seasonal variability of T. longicornis resembled (Fig. 3b) the pattern displayed by all nauplii, except at station 2. The seasonal cycle of hatching of C. hamatus (Fig. 3c) was most variable between stations.
–
2000
M
16
Hatching results (number of all nauplii that hatched from each sample in the first six screenings) were tested for significant differences between pairs of sampling dates. Values are differences of rank sums between sampling dates (**P < 0.05; *P < 0.1).
0.01
N6
9
N
D
J
F
2003
Hatching patterns
Sampling date
Fig. 2. Variability of hatching between sampling dates. Each column represents the sum of (&) N1, (&) T. longicornis, and (&) C. hamatus, which were found in the first six screenings in all samples collected at the respective sampling date.
samples taken per cruise were pooled. Sums ranged from 113 to 1582. Maxima were recorded in April, October and February and minima in June and December (Fig. 2). Friedman’s two-way analysis of variance by ranks indicated statistically significant differences between the seven sampling dates on the 99% confidence level [ 2R = 16.607; critical value for 2R with P = 0.01: 14.19 (Sachs, 1978)]. Wilcoxon’s matched pairs test showed differences between sampling dates April and June 2002 on the 95% confidence level as well as between April and December 2002 and between June and October 2002 on the 90% confidence level (Table IV). The seasonal pattern of hatching was similar in T. longicornis and C. hamatus. There were too few
(i) Short-term incubations In the majority of samples, numbers of nauplii were low in screening 1, but increased over the following 2–3 weeks. This was particularly true for those taken in March, April, October and February (Fig. 4a, b, e, g). Usually hatching continued during the entire experimental period. The pattern in the samples collected in June, August and December (Fig. 4c, d, f ), however, was quite different. Average numbers were comparatively high in screening 1, but declined thereafter. Towards the end of the incubation period values rose again gradually, especially in August and December samples. (ii) Long-term incubation Nauplii emerged from the samples collected in April at stations 2, 3 and 4 for the entire 52 weeks (Fig. 5b–d). Inter-specific differences were clearly expressed in the time necessary for 50% nauplii to hatch: T. longicornis needed less than half the time (mean 12 weeks, Fig. 6) required by
1087
j
JOURNAL OF PLANKTON RESEARCH
VOLUME
26
j
9
NUMBER
j
PAGES
1083–1093
j
2004
100
a
all nauplii
b
T. longicornis
c
C. hamatus
d
N1
Number of specimens in %
75
50
25
0 100
75
50
25
0
M
A
M
J
J A 2002
S
O
N
D
J F 2003
M
A
M
J
J A 2002
S
O
N
D
J F 2003
Sampling date
Fig. 3. Seasonal variability of hatching at individual stations. Values are the number of nauplii that hatched from a sample collected at a particular station (station 1, ; station 2, ~; station 3, &; station 4, *) and sampling date as percentage of the total number of nauplii that hatched from all samples that were collected at that station during this study. Only nauplii that were found in the first six screenings were considered. (a) All nauplii; (b) T. longicornis; (c) C. hamatus; (d) N1.
500 400
a
b
11-13 March 2002
c
22 April 2002
17 June 2002
300 200 100
Number of specimens
0 1
2
3
4
5
6
1
3
5
7
9
11
13
1
2
3
4
5
6
7
8
9
500 400
d
e
21 August 2002
f
15 October 2002
12-13 December 2002
300 200 100 0 1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
9 10 11 12 13
400
1
2
3
4
5
6
7
8
50
500
g
6 February 2003
h
all samples
40
300
30
200
20
100
10 0
0 1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
Screening
Fig. 4. Onset of hatching. (a–g) Number of nauplii that hatched per screening from all samples (April: screenings 1–13 only). Results are arranged according to sampling date (open circles, values of individual samples; black line, mean); (h) number of nauplii that hatched per screening from all 28 samples (open circles, values of individual samples; black line, mean; dashed line, mean based on secondary value axis).
1088
M. ENGEL AND H.-J. HIRCHE
j
HATCHING OF CALANOID COPEPOD RESTING EGGS
150
150
Number of specimens
a
b
Station 1
125
125
100
100
75
75
50
50
25
25
Station 2
0
0
1
13
26
39
52
1
13
26
39
52
250
150
c
d
Station 3
125
Station 4
200
100 150
75 100
50 50
25
0
0
1
13
26
39
52
1
13
26
39
52
Screening Fig. 5. Number of (&) N1, (&) T. longicornis, and (&) C. hamatus found per screening in samples that were taken at stations 1 (a), 2 (b), 3 (c), and 4 (d) on 22 April 2002.
Cumulative number of nauplii in %
100
75
50
25
0 0
13
26
39
52
beginning and decreased after 13 weeks . While T. longicornis seemed to have two phases of different hatching rates, C. hamatus displayed a third intermediate phase between weeks 13 and 25 characterized by an extremely low hatching rate. The rate was initially higher in T. longicornis but the situation changed after week 25. Only 27 Acartia spp. nauplii and 47 unidentified specimens were recorded from all four stations. The last Acartia spp. was found after week 51 and the last unidentified in week 47. In contrast, in the sample taken at station 1 hatching had already ceased after 38 weeks. The time required for 50% of the nauplii to emerge was also significantly shorter, 4 weeks for N1 and 6 weeks for T. longicornis and C. hamatus.
Screening
Fig. 6. Cumulative number of hatched nauplii in % (mean, range) for (*) T. longicornis and (~) C. hamatus across samples that were taken at stations 2, 3 and 4 on 22 April 2002.
C. hamatus (mean 27 weeks ). N1 stages reached this level after 14 weeks. Hatching rate was not constant, but changed with time in both species; it was higher in the
DISCUSSION Hatching experiments have frequently been used to demonstrate the existence of copepod resting eggs in marine sediments. Eggs were either extracted from the sediment (Marcus, 1989; Naess, 1991; Katajisto et al., 1998), or complete samples were incubated (Lindley,
1089
JOURNAL OF PLANKTON RESEARCH
j
VOLUME
1986; Hall and Burns, 2001). We applied the latter approach, which mirrored in situ conditions more closely than the incubation of extracted eggs would have done. Handling, however, considerably stirred up the samples, so that sediment stratification did not necessarily resemble the situation in the field. Nevertheless, the method should provide an estimate of the number of eggs likely to hatch in the event of only limited disturbance, comparable with normal tidal flushing (Lindley et al., 1998). The drawback of the incubation of complete samples is that it remains unclear how many viable resting eggs actually were buried in the sediment, as the experimental conditions may not have induced hatching in all eggs present. For example, nauplii hatched continually for one year from most samples collected in April, though hardly any were found to emerge from those that were taken at the equivalent stations in June. Hatching might have been suppressed by unfavourable conditions during the incubation. By a similar argument refractory diapause eggs may have dominated in the samples or sediments were depleted, if all resting eggs in the field hatched before June. However, numbers of nauplii found during the long-term incubations (maximum values were equivalent to 1.28 106 per m2) were in the same order of magnitude as direct egg counts in other studies (Guerrero and Rodriguez, 1998; Katajisto et al., 1998). We therefore assume that these results are a good approximation of the abundance of calanoid copepod resting eggs in the sediments of the German Bight, although hatching had not entirely ceased after 52 weeks. Even though six of the calanoid copepod species that occur regularly in the German Bight ( Johannsen et al., 1999) have been reported to lay resting eggs elsewhere [Acartia clausi (Kasahara et al., 1974), T. longicornis, C. hamatus, C. typicus, L. wollastoni (Lindley, 1986), Anomalocera patersoni (Ianora and Santella, 1991)], only nauplii of T. longicornis, C. hamatus and Acartia spp. were found in this study, along with a few specimens of an unidentified species. In the study area the percentage of spawning females and the egg production rate in T. longicornis are particularly low in November and December, but nevertheless reproduction occurs all year round (Halsband and Hirche, 2001). In the German Bight, the primary purpose of resting eggs may thus not be to secure the survival of this cold-temperate species (Halsband-Lenk et al., 2002) during the winter. Supposing T. longicornis produces morphologically distinct ‘non-hatching’ eggs predominantly in late spring/early summer, as described by Castellani and Lucas (Castellani and Lucas, 2003) for this species in the Irish Sea, the result may rather be to minimize mortality from predation and to prevent intraspecific competition. The decline in copepod numbers
26
j
NUMBER
9
j
PAGES
1083–1093
j
2004
near Helgoland in June has been attributed to predatory ctenophores like Pleurobrachia pileus (Greve and Reiners, 1980). Furthermore the phagocytic dinoflagellate Noctiluca miliaris, which peaks in the German Bight in July (200–300 cells L1) (Uhlig and Sahling, 1982), has been shown to feed on copepod eggs on a big scale (Daan, 1987). However, Daan noted that digestion time of these eggs by Noctiluca was very long. It is therefore reasonable to assume that the differences in egg shell thickness between subitaneous and ‘non-hatching’ eggs (Castellani and Lucas, 2003) should make the latter, more robust type, a less easily digestible food item. Diapause eggs from other copepod species have even been shown to survive ingestion unharmed (Marcus, 1984). Consequently, T. longicornis would benefit from the production of ‘non-hatching’ eggs at the peak of the reproductive season in May in at least three ways. First, it would minimize mortality of post-embryonic stages by decreasing the number of nauplii that hatch in times when predators are abundant. Second, the embryonic stages would be less easily digestible in times when predators feeding on copepod eggs are abundant. Thirdly, intra-specific competition for food is decreased in times when the phytoplankton concentration in the water column dramatically declines at the end of the spring bloom (Halsband and Hirche, 2001). Some of our results support this assumption. The hatching pattern visible in samples collected in June and August indicate that resting eggs, if present, may be refractory diapause eggs. Thus hardly any hatching occurs after a comparatively high number of nauplii, which possibly originated from non-quiescent subitaneous eggs, were found in the first screening. The presence of such eggs on the seafloor is particularly likely in June as egg production rates are high in T. longicornis and C. hamatus at this time of year (Halsband and Hirche, 2001). Substantial hatching in October might also indicate that resting eggs are an adaptation to unfavourable conditions in summer, rather than winter, but it is uncertain whether these results reflect in situ dynamics, as a new generation is usually not observed for T. longicornis in late autumn (H.-J. Hirche; unpublished data). However, as T. longicornis nauplii in the present study still hatched from sediment samples after an experimental period of one year, resting eggs may at least be capable of securing year after year survival. Additionally, maximum hatching rates in April indicate an important role in spring recruitment. Second most numerous to emerge from the samples were nauplii of C. hamatus. In subtropical waters resting eggs enable this species to survive adverse conditions in summer (Marcus and Lutz, 1994). In the temperate German Bight however, they are more likely to ensure
1090
M. ENGEL AND H.-J. HIRCHE
j
HATCHING OF CALANOID COPEPOD RESTING EGGS
Table V: Hatching results from long-term experiments performed in this study are compared with data from the literature (Lindley, 1990) This study
Lindley
Week 1–26
Week 27–52
Ratio
Month 0–6
Month 6–12
Ratio
T. longicornis (N2 and older)
1926
663
2.9:1
159
79
2:1
T. longicornis (N1–N6)
C. hamatus (N2 and older)
776
538
1.44:1
79
76
1:1
Centropages spp. (N1–N6)
N1
2305
895
2.57:1
For both studies, numbers of nauplii that hatched in the first and the second half of the one-year experimental period are given together with their ratio. Numbers of nauplii were pooled for four and eight samples in this and Lindley’s study, respectively.
the perpetuation of the species during the winter, as adults are absent from the plankton from December to March (Halsband and Hirche, 2001). But convincing evidence in support of this assumption is scarce in our results. The seasonal pattern of hatching displayed by C. hamatus in the experiments was similar to the one found in T. longicornis, although one would expect a comparatively larger peak in spring in a species that exclusively depends on recruitment from resting eggs at that time of the year. Moreover, pelagic populations of C. hamatus were observed to hibernate in other areas of the North Sea (Lindley and Hunt, 1989), and thus resting eggs may not be vital to the population in the study area, as planktonic stages may be advected to this region. However, based on the abundance of post-embryonic stages in the plankton (monthly means of T. longicornis near Helgoland are roughly ten times as high as in C. hamatus; W. Greve and F. Reiners, Hamburg, personal communications), the proportion of resting eggs in the sediment was considerably higher in C. hamatus than in T. longicornis. This militates in favour of their particular importance to this species. Significant inter-specific differences were also visible in long-term hatching. Unlike T. longicornis, the pattern in C. hamatus displayed an intermediate 3 months period characterized by an extremely reduced hatching rate. This resembles a refractory phase typical for diapause and was found to last as long as 6 months in diapause eggs of C. hamatus from the Gulf of Mexico (Chen and Marcus, 1997). But if the resting eggs present in the samples collected in April were spawned during the previous reproductive season (April–November), it is unlikely that they were still in need of time to complete diapause in August the following year. Long-term hatching also varied in other respects. Counts of C. hamatus nauplii decreased by only 30% in the second half of the experimental one-year period, while numbers of T. longicornis nauplii declined by 65%. Similar results were evident in a study by Lindley (Lindley, 1990). When he
incubated sediments from the English Channel twice as many T. longicornis emerged in the first 6 months of the experiment compared with the subsequent 6 months (Table V). In contrast, numbers of C. hamatus nauplii were almost identical in the two successive 6 month periods. These ratios are analogue to those found in our study and therefore support our findings. However, it is rather difficult to elucidate the exact mechanisms underlying the patterns observed, as it has neither been established whether in the North Sea resting eggs of both species include quiescent subitaneous and diapause eggs, nor what the seasonal cycle of the production of diapause eggs is like. The environmental triggers, which initiated hatching in the incubations, are also unknown, but our findings suggest that sample handling or sampling provided a signal that induced the eggs to commence hatching, as numbers of nauplii were often low in screening 1, but increased over the following two to three weeks. This was particularly true for those taken in March, April, October and February. Physical parameters like oxygen concentration and light have been identified as potential triggers for copepod eggs from different areas (Uye and Fleminger, 1976; Uye et al., 1979). In the present study, temperature remained constant in individual samples but the sediment surface was stirred up in every screening when bottles were refilled with seawater. Thus resting eggs were resuspended and exposed to new, possibly beneficial levels of environmental variables. Unlike Marcus and Lutz (Marcus and Lutz, 1998) we did not find a threshold temperature for hatching of C. hamatus resting eggs. Differences between stations 1 and stations 2, 3 and 4 in the long-term experiments may have been caused by the reduced longevity of diapause eggs at normoxia (Marcus and Lutz, 1998). Taking into account the differences in grain size distribution at the sampling sites such a situation is most likely to be encountered in the sediment at station 1.
1091
JOURNAL OF PLANKTON RESEARCH
j
VOLUME
If not hidden in the N1 group, no C. typicus nauplii emerged from the samples, although adults of this species are absent from the water column around Helgoland from February to August (Halsband and Hirche, 2001). Smith and Lane did not find evidence for diapause egg production in C. typicus either (Smith and Lane, 1987), but Lindley reported that some, though only very few, nauplii of this species emerged from sediment samples from the English Channel and southern North Sea (Lindley, 1990). The lack or scarcity of resting eggs and the absence of the species from the plankton for part of the year suggests that C. typicus is expatriated in the German Bight and depends on advection with Atlantic water, as already hypothesized by Krause et al. (Krause et al., 1995). CPR (continuous plankton recorder) data, however, suggests that the species overwinters in the German and the Southern Bights (Lindley and Reid, 2002). Nauplii of Acartia spp. could not be determined to species level. Several species of this genus occur around Helgoland, but A. clausi is most abundant (Krause et al., 1995) and present in the plankton all year round. During the winter 1995/6 it did not reproduce from October to January (Halsband and Hirche, 2001), which makes the species a candidate for resting eggs as an overwintering strategy. A. clausi resting eggs have indeed been reported from the Pacific (Kasahara et al., 1974; Uye and Fleminger, 1976; Marcus, 1990), but appear to be particularly vital to populations living in warmer waters (T > 22 C), where they guarantee survival during the summer, when the species is absent from the plankton (Uye, 1985). They also occur in areas of the Pacific, where water temperature is lower (even below 0 C during the winter), but are probably of less importance, as planktonic populations are perennial in these regions [see Koyama, 1975 in (Uye, 1985); (Uye, 1982)]. In the Atlantic clear evidence is missing. As A. clausi is as abundant in the German Bight as T. longicornis (W. Greve and F. Reiners, Hamburg, personal communications), corresponding numbers of resting eggs are to be expected in sediments. Therefore the small portion of Acartia nauplii found here are more likely to originate from A. bifilosa and/or A. tonsa, which are found around Helgoland and are known to have resting eggs (Uye and Fleminger, 1976; Viitasalo, 1992). Labidocera wollastoni and A. patersoni are both rare in the study area, monthly means are
